AI agents are becoming part of the crypto workflow. They can research wallets, inspect transactions, monitor token activity, and automate on-chain analysis.
However, raw blockchain data is difficult for an agent to interpret on its own. Solana activity is distributed across wallets, token accounts, programs, and individual instructions. A single transaction may include transfers, swaps, fees, balance changes, and interactions with several programs.
To work effectively, agents need structured context such as transaction types, token metadata, account relationships, DeFi activity, and decoded program interactions.
Solscan supports these workflows through three layers:
- Solscan MCP for structured data access
- Solscan CLI for command-line queries and automation
- Solscan Agent Skills for reusable workflow instructions
Each serves a different purpose while using the same underlying goal: making Solana data easier for agents and developers to work with.
1. Solscan MCP: Data Access for AI Models
Solscan MCP connects AI models and agentic applications with structured Solana data.
MCP stands for Model Context Protocol, a standard that allows AI systems to access external tools and data sources through a consistent interface. Instead of working directly with raw RPC responses, an agent can retrieve indexed information such as wallet activity, token balances, swaps, market data, decoded transactions, and program interactions.
For example, when a user asks an agent to summarize a wallet’s recent activity, the agent may need to review transfers, swaps, balance changes, and program usage. Solscan MCP provides the structured data required to generate that summary.
MCP is most suitable for conversational agents and AI applications that need Solana data during a live workflow.
Learn more: Solscan MCP document.
2. Solscan CLI: Queries and Automation for Developers
Solscan CLI provides terminal-based access to Solscan data.
It supports more than 55 actions and returns results in both human-readable and JSON formats. Developers can use it to inspect accounts, retrieve token data, review transactions, monitor wallet activity, and connect Solscan data with scripts or backend systems.
Human-readable output is useful for direct review, while JSON output is better suited to automated workflows. A developer could, for example, run scheduled wallet checks, filter recent activity, store the results, and pass them into another monitoring or reporting process.
The CLI can also be used by agents with terminal access as part of a larger automated workflow.
It is most suitable for developers, analysts, scripts, scheduled jobs, and internal tools.
Learn more: Solscan CLI document.
3. Solscan Agent Skills: Reusable Workflow Logic
Solscan Agent Skills provide reusable instructions for completing Solana-related tasks.
A data tool tells an agent what information it can access. A Skill tells the agent how to use that information for a specific workflow.
For example, a wallet analysis Skill may instruct an agent to retrieve recent activity, separate transfers from DeFi actions, identify the main tokens involved, and produce a consistent summary. This prevents developers from rebuilding the same workflow logic for every implementation.
Skills are useful for tasks such as wallet analysis, transaction explanation, token research, swap review, and multi-account comparison.
They are most suitable for teams building agents that need consistent and repeatable Solana workflows.
Learn more: Solscan Agent Skills document.
How the Three Layers Differ
The distinction between the three layers is straightforward:
| Layer | Main role |
|---|---|
| Solscan MCP | Gives AI models access to structured Solana data |
| Solscan CLI | Lets developers query and automate Solscan data through a terminal |
| Solscan Agent Skills | Defines reusable instructions for completing Solana workflows |
They can be used independently or combined.
A conversational agent may use MCP to retrieve wallet activity. A developer may use the CLI for scheduled monitoring. A research agent may use Skills to define how retrieved data should be analyzed and presented.
In a combined workflow, MCP provides the data, Skills define the analysis process, and CLI supports automation.
Conclusion
AI agents need more than raw blockchain records. They need structured context that explains wallet activity, token movements, transactions, programs, and DeFi actions.
Solscan supports this through MCP for data access, CLI for automation, and Agent Skills for reusable workflow logic.
Together, these tools provide a practical foundation for building Solana research assistants, monitoring systems, transaction analysis tools, and other agent-powered applications.